1. Field of the Invention
This invention relates to photoplethysmographics. More specifically, this invention relates to component-amplitude-division multiplexing and demultiplexing of signals for infrared and red absorption of blood.
2. Description of Related Art
It is well known in the art to collect photoplethysmographic data simultaneously at a plurality of energy wavelengths. For example, blood oxygen concentration may be measured by determining absorption by a patient's tissues on infrared and red light; the degree of absorption is typically different for these two wavelengths. Infrared and red light are emitted into the patient's tissues (e.g., by infrared and red LEDs) and the total energy received to be detected by a single detector (e.g., a photodiode). However, one problem is that the signal produced by the detector must be processed to separate the infrared and portions from each other.
One method of the prior art is shown in U.S. Pat. No. 4,407,290. Time-division multiplexing is used to alternately switch on the infrared and red emitters, at a frequency greater than the patient's pulse rate. The detector signal is then separated into infrared and red portions by sampling in synchrony with the on/off switching of the infrared and red emitters.
While this method successfully separates the infrared and red portions, it generally requires that sampling the detector signal must be synchronized with the on/off switching of the infrared and red emitters. It is also difficult while using this method to compensate for noise sources such as ambient light and electromagnetic interference.
A second method of the prior art is shown in U.S. Pat. No. 4,800,885. The infrared and red emitters are driven at two different frequencies. The detector signal is then separated into infrared and red portions by filtering at those two different frequencies.
While this method successfully separates the infrared and red portions, the method described in the patent requires demultiplexing signals which are phase-synchronized with the multiplexing frequencies, and produces a higher power output than the time-division multiplexing method. Also, while this method may avoid noise sources at predetermined and known frequencies, it is difficult to compensate for noise sources which were not known before the multiplexing frequencies were chosen.